As I am trying positions for the updated PSU in my monoblocks, I find that the transformer and switch circuit fill the space near the opposite side from the M2x board. (Monoblocks) this moves the SLB into the middle of the case. This keeps all the AC on the far side.
My question is about the outboard mosfets. Can they be mounted on the steel plate on the bottom of the 4u deluxe case. Or longer fly leads to reach the heat sink on the side opposite the AMP board? Not sure how much dissipation is needed. Hmm, possibly some aluminum heat sinks mounted to the base close to the SLB...
My question is about the outboard mosfets. Can they be mounted on the steel plate on the bottom of the 4u deluxe case. Or longer fly leads to reach the heat sink on the side opposite the AMP board? Not sure how much dissipation is needed. Hmm, possibly some aluminum heat sinks mounted to the base close to the SLB...
I have one that I have never run across... I've built about 8 of these supplies and never had one give me an issue. I have one side that is almost 2v high. The transformer is fine, but the output on one side is about 2v high. Also the variable resistor does nothing on that side.
I haven't done any troubleshooting yet... I thought I would drop a note here and get your thoughts before I did so.
The pos rail unloaded is the one that is high.
Thanks,
JT
I haven't done any troubleshooting yet... I thought I would drop a note here and get your thoughts before I did so.
The pos rail unloaded is the one that is high.
Thanks,
JT
I never noticed that with the SLB, but I have had other psu’s that exhibit odd behavior when they are ‘free wheeling’ without a load.
Also, without current being drawn, the SLB trimpots have little to no effect when turned.
Give a close look over the board to see if you spot an incorrect part value or cold solder joint. If all looks good, try putting a resistor load on the outputs and fire her up! Wear your face and tactical gear just in case
Ps: Don’t forget to connect the BJT’s, otherwise all the current passes through a .5w resistor and POOF!!
Troubleshoot
On the trimpot, yes, but you should be able to see 1/2-3/4 of a volt or so. I see that on the good side, but point taken. We both agree that the SLM isn't likely to do this without something being amiss even unloaded.
So here is what I'm thinking... I dumped off a couple copies of the print. I'm going to pull the board out of the chassis, so I can get at the bottom of the beast and do some node testing on it. I'll get some voltages and post them with any thoughts I may have on it.
Besides, if I don't test some nodes X is going to come in a bust by tests.
I never noticed that with the SLB, but I have had other psu’s that exhibit odd behavior when they are ‘free wheeling’ without a load.
Also, without current being drawn, the SLB trimpots have little to no effect when turned.
Give a close look over the board to see if you spot an incorrect part value or cold solder joint. If all looks good, try putting a resistor load on the outputs and fire her up! Wear your face and tactical gear just in case
Ps: Don’t forget to connect the BJT’s, otherwise all the current passes through a .5w resistor and POOF!!
On the trimpot, yes, but you should be able to see 1/2-3/4 of a volt or so. I see that on the good side, but point taken. We both agree that the SLM isn't likely to do this without something being amiss even unloaded.
So here is what I'm thinking... I dumped off a couple copies of the print. I'm going to pull the board out of the chassis, so I can get at the bottom of the beast and do some node testing on it. I'll get some voltages and post them with any thoughts I may have on it.
Besides, if I don't test some nodes X is going to come in a bust by tests.
Testing
Thanks for the input Don, but I think I'm going to safe route and do some testing before dropping a load on it and hoping something doesn't smoke.
@Vunce
I'm thinking a 25W potentiometer you could set to 8ohm to look like a speaker and 25w should be safe. Am I on the right track? I did similar with a 3w resistor and poof.
Don
Thanks for the input Don, but I think I'm going to safe route and do some testing before dropping a load on it and hoping something doesn't smoke.
The BJTs burn off 2.5A x 3v or 7.5w each. The steel panel on the bottom deck should be fine. Make sure you use a good insulator. With sheet steel, be careful not to have a sharp burr like through the insulator and you have a dead short from rails to ground.
@Vunce
I'm thinking a 25W potentiometer you could set to 8ohm to look like a speaker and 25w should be safe. Am I on the right track? I did similar with a 3w resistor and poof.
Don
The dissipation from a typical 25w Class A amp is circa 60w. You will need a 100w power resistor. One of those big wirewound ceramic jobs. Or an EBG flat bulk metal one bolted to a heatsink. Even a 25w resistor would burn up if given 60w to dissipate. Or 4x 25w ones in series parallel.
You can calculate value needed assuming 2.5A bias current (both channels). +/-24 v is 48v. R=V/I or 48v/2.5A=19.2ohms. 22ohms would be close enough. That’s the test load you need.
On the trimpot, yes, but you should be able to see 1/2-3/4 of a volt or so. I see that on the good side, but point taken. We both agree that the SLM isn't likely to do this without something being amiss even unloaded.
So here is what I'm thinking... I dumped off a couple copies of the print. I'm going to pull the board out of the chassis, so I can get at the bottom of the beast and do some node testing on it. I'll get some voltages and post them with any thoughts I may have on it.
Besides, if I don't test some nodes X is going to come in a bust by tests.
2v too high sounds like the cap Mx is not working and feeding straight through. Might be dead BJT (shorted so passes whatever through) or cold solder joint on one of the peripherals around the cap Mx. or blown resistor that’s open circuit somewhere.
Very nice devices but the input/output setting is not so clear (at least for noobs, except if missed some important info).
From the 1st post:
"Update May 5, 2019: Verification build test with 4.4A and 35.4v out and 3.1v drop gives about 1mVrms ripple with the SLB (with R17/18 replaced by jumper)"
--> Shall we consider that the drop through the SLB is 3.1V whatever the input voltage?
Maybe a comprehensive table 'tranformer voltage -> expected output voltage from the SLB' could help.
From the 1st post:
"Update May 5, 2019: Verification build test with 4.4A and 35.4v out and 3.1v drop gives about 1mVrms ripple with the SLB (with R17/18 replaced by jumper)"
--> Shall we consider that the drop through the SLB is 3.1V whatever the input voltage?
Maybe a comprehensive table 'tranformer voltage -> expected output voltage from the SLB' could help.
It depends on the brand of transformer. Antek’s are inexpensive and available in 2v increments. I typically size them up because they have higher drop under load vs other brands. The drop also depends on current draw.
Here is my usual approach to estimate the transformer dropout:
1. Work backwards from desired voltage after the SLB under nominal Class A load. For example, most Pass 25w amps are spec’d to have +/-24v at 1.25A. Decide if you will use one SLB for both channels or dual monoblock. Single SLB would imply 2.5A total current.
2. Size transformer ~3x required VA rating. 2.5A x 48v is 122w so 300VA is not enough. Next size up is 400VA.
3. Assume 3-4v drop or sag in nominal voltage rating of transformer. Assume 3v drop in SLB for ripple regulation. So 6v total dropout. Add 6v to 24v and we get 30vdc. Divide 30 by 1.41 to get the AC voltage rating of trafo. That’s 21.4vac. With Antek always size UP. So between a 20vac and 22vac, choose 22vac.
So there is the required trafo: 400VA and 22vac for single trafo.
For other brands like Toroidy, Avel Lindbergh, etc. it depends but I find they don’t have the 3-4v sag. Maybe 1-2v only. The Antek ones have an accurate LTSpice model on the website. You can use that to calculate the sag better.
Here is my usual approach to estimate the transformer dropout:
1. Work backwards from desired voltage after the SLB under nominal Class A load. For example, most Pass 25w amps are spec’d to have +/-24v at 1.25A. Decide if you will use one SLB for both channels or dual monoblock. Single SLB would imply 2.5A total current.
2. Size transformer ~3x required VA rating. 2.5A x 48v is 122w so 300VA is not enough. Next size up is 400VA.
3. Assume 3-4v drop or sag in nominal voltage rating of transformer. Assume 3v drop in SLB for ripple regulation. So 6v total dropout. Add 6v to 24v and we get 30vdc. Divide 30 by 1.41 to get the AC voltage rating of trafo. That’s 21.4vac. With Antek always size UP. So between a 20vac and 22vac, choose 22vac.
So there is the required trafo: 400VA and 22vac for single trafo.
For other brands like Toroidy, Avel Lindbergh, etc. it depends but I find they don’t have the 3-4v sag. Maybe 1-2v only. The Antek ones have an accurate LTSpice model on the website. You can use that to calculate the sag better.
Last edited: